Month: September 2023

The theme of Knobcon 11 — which has previously been such things as “Jack to the Future”, “Number Six” (The Prisoner), and “Knobtoberfest”, was utterly inevitable:

Tired from the drive and a little overwhelmed by the ridiculously noisy lobby (it has the worst acoustics ever, so it sounded like a World Cup match even though there was plenty of space and I maintained about 60 feet of social distancing until I actually checked in), I waited until about 4:30 to go down and do the con registration. There was quite a line. Maybe 1 in 5 people wearing a mask, which is more than average in the general public I think. I’m definitely going to stick with that in any crowded area given that Covid is on the rise again and the new booster won’t be available for another couple of weeks.

Gotta go back tomorrow to pick up my t-shirt, but I got my badge and bag o’ tchotchkes. This time there was no 9-pound dead tree Sweetwater catalog, thank goodness, and only a handful of cards and stickers and paper-thin fridge magnets I tossed out. Here’s the better and/or more bizarre stuff:

• An actual, quality wooden coaster from Oberheim. Nice!
• A decent, longish patch cable from SynthCube.
• A pin, sticker, and 15% discount card for online orders from SSF.
• I kept the EQD sticker because they have one of the best logos in the business. 🙂
• Perfect Circuit’s thingy includes a little screwdriver. I probably already have one, and I’ve probably lost another one somewhere. But hey.
• A squishy stress ball from Rogan. (It is not representative of the rubber they use for their soft-touch knobs.) I am amused that an actual knob company is sponsoring Knobcon now. But they do make my favorite knobs (as found on Make Noise, IME/The Harvestman, and Mutable Instruments by default, and I’ve put them on several other modules too).
• A PCB (circuit board / non-metallic panel) business card from Infinite Machinery, with what appears to be a surface-mount 0-ohm resistor on it. (If it actually is a zero ohm resistor, it’s time to alert the press that they’ve invented room temperature superconductors…) Maybe I can tape it over the very poorly located power switch on my computer that dogs have been known to turn off with their noses or butts.
• An aluminum circle from MetalPhoto. I don’t know what I’m going to do with an aluminum circle. Last time it was a ruler, which is a lot more practical.

In about an hour and a half, the festivities (live music) start up. Right now I’m just chilling in my hotel room and thinking about getting into my snack stash rather than bothering with real dinner.

---

Yesterday during my work hours, we had an internet outage. When it became apparent it was going to be down for a while, I signed into Teams on my phone’s slow and unreliable 4G connection just in case, and fired up the synths. I believe now I just have one, maybe two more tracks to record to round out the album.

Entonal Studio’s MIDI support works with Univer Inter, so I was able to play Strega in a subset of 18edo from the Seaboard. There was a little bit of wobbling at the attacks of some notes, but I lived with it for the purposes of that one track as it gave some additional character. This is masked when you’re using pressure control rather than an envelope with a relatively fast attack — but since my internet was down I was stuck with the settings I’d last uploaded to the Uni.

Pitch being slower than gates is a thing that happens with Eurorack sequencing at times, which can be solved with a very short gate delay. With MIDI normally you get the pitch and gate in a single combined message — but Entonal Studio is also sending pitch bend messages to adjust the tuning, so between it and the MIDI-to-CV conversion there’s an opportunity for things to get weird. But I can fix it easily enough with short gate delays using Stages or Teletype, no problemo.

So obviously, this is The Season That I Got Into Microtonal Stuff (STIGIMS).

I wound up grabbing Entonal Studio, a plugin that manages microtuning. I really wish I’d picked it up earlier — it’s what I should have been using all along.

Bitwig Micro-Pitch has two modes:

• 12-Note mode, where you can tune each keyboard key and resize the octave. All units are in 12TET semitones, aka “dollars” (100 cents) with no conversion from ratios, EDO etc. If you want a scale with less than 12 notes you’ll need to copy and paste values; if you want one with more than 12, too bad.
• EDO mode, where you set just the interval size and number of divisions. You can’t skip divisions, so if you want a nice diatonic scale with 19edo tuning, you’ll have to work out the conversion elsewhere (e.g. Scale Workshop) and enter it into 12-Note mode instead.
• Both of them give you control over the root note and fine tuning. There’s also an intriguing “wet/dry mix” that interpolates between 12TET and your specified tuning. (And yes, that is extremely weird with EDO mode.)
• There’s a decent set of useful presets.
• It works via MPE, effectively sending pitch bend messages for each note. It works poorly with non-MPE synths.

So, while you can do some cool things with it, it’s got its limits.

And then there’s Entonal Studio. You can have any number of notes in your scale, and you can specify each one by ratios, cents, divisions, etc. You can just drag pitches around the circle to adjust them if you like, and optionally snap to ratios or equal divisions. There are very handy visualization tools, which help not just in constructing a scale but when playing it. You can edit the mapping any way you want. You can see a table of MIDI note numbers to names, Hertz, cents, the expression that created it, and its ratio relative to a root frequency. There’s also a Lattice view, which gets into aspects of theory that I skimmed over in the Gann book.

It supports not just MPE, but monophonic MIDI (using pitch bend), multichannel MIDI, and MTS-ESP (a new quasi-standard not yet widely supported, but Plogue OPS7 can use it). It can host a plugin within itself (which tends to work better for non-MPE synths). It can import and export Scala and XML tuning files.

What’s it missing? Hmm…

• The ability to compare two scales. It’d be nice to see how they align with each other, whether you’re checking for general consonance or because you want to play different scales simultaneously. (Comparing to 12TET isn’t hard if you just look at the cents values, but seeing it visually would still be nice.) I’ll probably write to the developer to ask about this!
• A tool for approximating by ratios or divisions, like Scale Workshop does. But you can sort of do this interactively with the snap feature, and you can always export a Scala file, mess around in SW and re-import, which gets you a lot of other nerdy stuff.
• This is more of a pie in the sky thing, but after reading up on dissonance curves in the Sethares book, I kind of want that integrated into scale creation software. It’d be very cool to enter a few partials and their amplitudes (or even take an FFT of a sound for you), and have it generate a dissonance curve and have it fit possible ratios or EDO scales. This would be major new feature territory or even a separate addon or product, but I might ask about it too.

Having finished one book on tuning systems, I am now on Tuning, Timbre, Spectrum, Scale by William Sethares. Overall, it explains how consonance and dissonance work, and how if you’re interested in consonance, you have to match your tuning/scale to your timbre or vice versa.

The quick version:

Most instruments in Western music have a mostly harmonic spectrum, because of the physics. Strings are held in place at both ends and can’t vibrate, so any waves must be exactly that strength length or must divide the length equally (so 1/2, 1/3, 1/4 etc.) Air columns in a wind/brass instrument, organ pipes or the human voice behave similarly. With any acoustic instrument there is generally some additional mechanical noise, and factors like string tension which can warp things a bit, but overall it’s pretty close.

The harmonic spectrum is why small ratios for intervals work so well — going up to 2/1 or 3/2 or 4/3 or 5/4 etc., you’ve got partials that match up with the root’s partials. The higher the ratios, the higher up the spectrum (and thus weaker) those matches are — and more importantly, some of the non-matching partials wind up in a zone where they clash with each other.

But percussive instruments are inherently inharmonic. You can carve a wooden arch in the back of a xylophone bar, shave material off a church bell in strategic locations etc. to try to to reduce the dissonance either by weakening some vibration modes or bending them into better frequencies, but you can’t completely fix it. And of course synths can create inharmonic tones using FM, additive synthesis, or a frequency shifter.

The book points out that these nice small ratios, even octaves, can be dissonant when the timbre is inharmonic. But you can find a tuning/scale that fits the spectrum. This is what Java and Bali do with their gamelan scales, tailored to fit the spectra of the individual instruments.

Conversely, you can find spectra that will complement exotic scales such as 10edo, which is kind of a problem for harmonic instruments.

With the majority of my 18edo-based album already recorded, I wonder if I can find a Tension setting for Odessa that results in consonant neutral thirds. Mostly I’ve just played what felt right and interesting, but I do tend toward minimal harmony and not following the theoretical rules about chord progressions, voice leading etc. anyway.

Since I’ve been deep in this 18edo album project, I thought it’d be a good time to learn more about microtonality and tuning systems in general. So I grabbed Kyle Gann’s The Arithmetic of Listening: Tuning Theory and History for the Impractical Musician off my wishlist and dug in.

I love that title BTW. I certainly can be an impractical musician at times. Not quite as deep into that as some of course, but I do feel that the most efficient path from A to B isn’t always the most creatively inspiring one.

The material in this book more or less divides into two main areas. The first is the quest for a perfect tuning system, which is unfortunately not mathematically possible.

The most consonant sounds — the “sweetest”, purest, least clashing — are, as the Mesoptamians and Greeks proved, the ones with the simplest integer ratios. 2:1 (called the octave for dumb historical reasons), 3:2 (the fifth) and 4:3 (the fourth). It’s downhill from there, but it’s the major and minor thirds that make Western chords what they are.

The problem is, if you insist on using 3:2 as a fifth, the “circle of fifths” used to build a scale is really a spiral. 3/2 to the 12th power is 129.746, which is not a power of 2. Over the centuries, various strategies were invented to cope with this dilemma. Every one of them has its problems, from “you cannot play in that key on this instrument” to “G# is not the same as Ab so I need more than 12 keys for my 12-tone scale” to “absolutely none of these intervals is really in tune.”

Meanwhile in the world of fretted stringed instruments, they never had a choice but to use equal temperament. All of the semitones have the same ratio between each other, though it’s not based on “nice” ratios. Historically when you placed frets, you’d measure 1/18th of the length of the string between the previous fret and the bridge. (This number is more accurate but awkward today.) When you got to the 12 fret you’d be very close to an octave. And coincidentally, the fifth and fourth were pretty close to being right too… it’s just that the thirds and sixths are kind of broken. But we got used to it.

Last night, I was playing pure tones using Bitwig Micro-Pitch to try different intonations and temperaments. This is super easy compared to retuning a piano or building a different organ, of course. In a high enough register it’s shockingly easy to hear these conflicts in tuning that we just take for granted. But then, they can be masked by more complex timbres and arrangements, and the way we compose and play music in the first place.

All of that was not the reason I got into this subject though.

The other aspect of tuning theory is: what do you do if you want to break out of that system and make something that sounds more exotic, eerie, unfamiliar?

There’s a lot of fancy theory here too. Some composers have built scales using other ratio theories. But one of the simpler things is to explore is other equal divisions of the octave (edo) besides 12, or even another basis besides the octave (such as the tritave, the 3:2 ratio). Thus my playing around with 7edo at times, and 18edo in this project. And mostly that started out of curiosity, back when I had the ER-301 Sound Computer in my rack.

In some of those macrotonal scales you just don’t get to play “chords” at all and all of the intervals sound odd. And given the kind of music I make, that’s OK. I don’t really think about Western harmony concepts in my music, though they may happen by accident (I hope that pun doesn’t fall flat…)

In my current project I’m not being super strict about this. I have too many different sources, drones tuned by ear, harmonic series that are inherent a part of the timbre, frequency shifting that throws everything off. But the compositional basis of it is still 18edo.

---

I’m kind of putting together my thoughts on tuning, scales and quantization options and how to take advantage of them in my hybrid Eurorack/DAW setup. I was planning on writing that up here tonight, but it’s starting to get late and I’m still rearranging the blocks in my head. I’ve tended to do things intuitively without theorizing much, but it’s possible that learning this (and maybe more importantly, finally hearing the differences in different intonations/temperaments) will lead me somewhere. Maybe it’ll just be the occasional dip into a just-intoned triad left bone-dry for the shock factor of zero beating.